CHDyshli / HrSegNet4CrackSegmentation

Real-time High-Resolution Neural Network with Semantic Guidance for Crack Segmentation
Apache License 2.0
46 stars 0 forks source link

Need to understand the output #13

Open Krishna1Gupta opened 3 months ago

Krishna1Gupta commented 3 months ago

I have used the HrSegNetb 16 and HrSegNetb 48 model weights for doing the inference on the below image, I got the following results. I can not able to infer these results. I have tried other images as well as other HrSegNet Models, but I am still getting similar two channel images. I will be highly grateful if someone can explain these.

Input_Image crack

Infered Image from HrSegNetb 16 infer_2

Infered Image from HrSegNetb 48 infer_1